213 réponses à ce sujet

[RigAudio]

Cra5y Pineapple wrote...
Would suggest modifying the turian speech. A bit too normal.

Yeah, I'm still modifying that. My first pass was too dumb.

Highdragonslayer wrote...
Do you think can do an overmind one
too?

I think I could. Sounds like a simple process, but it could turn into a needlessly complicated one, lol

[Ohem85]

(Sorry for bumping the thread, but I have yet to find a more recent)



As I like to play with sounds from time to time with Reaper Audio Software (although not too experienced), I wouldn't mind what settings, plugins etc you're using that I could use myself (if I'm not asking too much that is).



I really love the results you made with Turian, Drell and Sovereign.

[problemaker]

OMG! I'd love to have a program that could put my voice on the mic sound like sovereign's.

I'd pay for that! xD

Awsome sound results rig, ur really almost there with sovereign!

[PsyrenY]

problemaker wrote...

OMG! I'd love to have a program that could put my voice on the mic sound like sovereign's.

There's a number of free audio editors you can play around with. Try Audacity.

[Greed1914]

All I know is that I'd talk to Sovereign every day. That guy sounded awesome.

[Torley]

This thread is a lot of fun, thanks RigAudio for starting this out — and everyone who's contributed to research. I never thought I'd laugh so hard at something I'd never imagine a Drell saying. But that's generally true for having notable voices speak "out of character".

Inspired... I was feeling dizzy and had a tummyache the other day, then had a go at doing the Geth/Legion voice in Ableton Live 8 (settings are shown)! Ah, those beloved lampheads. Like Sovereign and some others, seems like it just takes the right vocoder settings to get it done. It'll depend on your original timbre, of course, but the end result is still recognizable.

[Wolfy2449]

Nice

Drell seemed quite good

Reapers was funny but thats how the voice is xD

I dunno about turians though, its just seemed like a human light voice with some small differencies, got no idea on suggestions though, not any experience with audio stuff



And yes we really need a legion-ify option xD

[TuringPoint]

I would be more interested in knowing how to do this myself than sending it off for someone else to have all the fun

Am I really the only one who thinks the Geth filter is pretty good?

It's worth noting that Bioware typically had a handful of people doing one race or another.  I'm pretty sure there was one guy with the standard turian voice.

Modifié par Alocormin, 18 novembre 2010 - 02:54 .

[Kitedtk]

I tried to search the forums for this, but couldn't find anything.

Does anyone have a working link of his speech files? Or anything like it?

What I would like is some help on how to access his speech files. I want to make my computer startup sound be his entire speech. I can combine the speech files etc. myself, but I need access to the actual files and am at a loss as to how to do this.

I tried to find the files, any information on extracting them, guides, just about anything, but to no avail. It would be absolutely awesome if someone was willing to help me out on this or atleast point me in the right direction.

In case someone is willing to help and I do get it all done, I want to share this with other people. Would be an awesome task to complete and share the benefits with the rest of the community.

His speech was probably the final thing that absolutely addicted me to the saga.

Modifié par Kitedtk, 22 novembre 2010 - 04:04 .

[Hi Tehk]

RigAudio, I love you for taking the time to show the steps on how to design voices like ME2. Im seriously obsessed with me sounding like a Drell or turian. Im on FL all night not even making music anymore. Just writing stuff and recording it XD. I wanna do a dubbed comic so bad now!

[Amdor]

RigAudio wrote...

Drayvenn wrote...

Please keep up the good work.
No love for Quarians yet?

Forgot about them, actually. Yeah, they should be pretty simple to replicate. I'll make a demo in a few days.

Hmm no updates made me a Sad Hanar

[Ramirez Wolfen]

You'll Turian-ify my vvoice? How do you do that? Do I have to send you a sample of my voice?

[sarahN7]

Ok, seriously? Why are all these old threads getting resurrected?

[Computer_God91]

sarahN7 wrote...

Ok, seriously? Why are all these old threads getting resurrected?

ZOMBIES!!!!!

Seriously though, does it matter? You aren't helping keeping them dead by posting in them yourself.

[Ramirez Wolfen]

sarahN7 wrote...

Ok, seriously? Why are all these old threads getting resurrected?

Does it matter? If you don't like the thread, why post in it?

[praetor_alpha]

I'm playing around turian-ifiying my voice in Audacity. If I pitch it down 6 semitones, I sound precipitously close to Saren. I wanted more of a pedestrian turian, rather than pure evil. Still toying around. Might play around with changing the speed of one of the tracks by .1%.

Just tried upping the original voice by 3 semitones, dropping the effect track, now I'm a salarian!

Modifié par praetor_alpha, 10 février 2011 - 05:05 .

[Ramirez Wolfen]

What kind of mic are you using?

[praetor_alpha]

Ramirez Wolfen wrote...

What kind of mic are you using?

A noise cancelling mic I got from radio shack about 10 years ago. Rather bland looking desktop mic with with a 12" boom.

edit: I think I'm on to something.
http://soundcloud.co...or_alpha/turian
http://soundcloud.co..._alpha/salarian

Modifié par praetor_alpha, 10 février 2011 - 11:29 .

[Taalii]

What exactly did you do?  Trying to replicate the Sovereign filter in MorphVox for use in voice chat, but I'm not too good with this stuff and cannot tell what you used.

[night_raven26]

 To make you sound like a krogan, speak with a deep voice and change pitch by -35 to -25 percent.

[DeathsHands5]

Yeah, I got a handful of audio programs around, but I'm not too certain on replicating Sovereign. A guide for that would be badass.

Modifié par DeathsHands5, 19 mai 2011 - 02:26 .

[night_raven26]

 This is how I tried the turian voice in Audacity: http://soundcloud.co...-voice-act-test

1. Recorded the sound (decreased -2.25 semitones because my voice is high)

2. Duplicated it

3. Took the second track, lowered the pitch for 6 semitones, then applied low pass filter with a cutoff frequency of 200 Hz

4. Saved the track (as testing)

5. Opened it again (the tracks merged)

6. Duplicated it

7. Took the second track and changed speed to 0.1 percent (that creates the flanging effect which can be heard in turian voiced)

8. Lowered the track volume to -6 db

9. That's it (try playing with it a bit)

[Enwinor]

Another revival...The real secret to an authentic Sovereign voice is speaking in sort of a quiet, throaty growl, but keeping your voice even and calm. It gives it that booming bass sound that makes it so menacing once you put in the effects. I have some examples saved in Audacity, but no links as of now. My impression isn't perfect, but it sounds very Reaper-y. For Harbinger, just use a lower pitch-shift on a deep voice and then apply the other effects. Delay for the mechanical "bite" in the Reaper voice should be about .03-.04 seconds.

Modifié par Enwinor, 29 mars 2012 - 09:34 .

[Merovign]

I wanted to resurrect this before and didn't - BUT...

The mising ingredient seems to be the process for Sovereign's voice, which no one seems to have written down. I did, after some experimentation.

A good mic is a good idea, and a windscreen. Sibilance gets nasty with this processing (s-sounds).

Also what enwinor said above. Go for deep but don't strain, you get deep from the processing. Evenness matters more.

I used Sony Sound Forge, but almost every function should be available in your editing software.

I'm assuming some basic audio knowledge but I'll try to give enough info for an amateur to do it.

Lastly if *any* feature in your editing software has a live preview - USE IT. The result will be much faster and closer to what you want.

1. Record. See above.

2. De-esser. Reduces artifacts of sibilance. In SF it's under the Multi-Band Dynamics plugin included, some other programs you will need to get a plugin, there should be free ones on the net for more popular programs.

3. Adjust your volume. You want audible but not too loud, some of these processes amplify. You don't want to create clipping.

4. Sound Forge has a built-in feature called the "Wave Hammer" - it combines certain frequency boosts, bass boost (mild) and compression. In a way it's kind of like auto-tuning. If you don't have anything like this, it's not really essential, it just makes a poor mic sample sound better.

5. Flange - use a flange effect. I'm not quoting specific numbers because it will depend on what you started with to some extent, also start small and be prepared to apply a filter repeatedly, that actually helps create a more artificial, ethereal sound. I used a rate of 0.023 Hz and a depth of 76%, but I may do more than one pass and alter settings.

6. Pitch shift - Sovereign is not actually *that* low, but there is a bass boost that makes it seem so. Try dropping a few semitones to start. Live preview *really* helps here. You *might* want to do this after the Chorus effect. If there's an accuracy setting move it to low accuracy - again, sounds more synthetic.

7. Chorus. Start with 3 voices, attenuate high frequencies if possible (above 4k-5k), reduce the delay to something very short, say 30ms. If you add 3 more voices, invert the phase if you have the option to.

8. If you have a graphical EQ filter, esp. chart-based, drop out a narrow window of sound in the midbass range (160-600Hz) - you can move that range around a bit. I notice playing with the original Sovereign files there seemed to be a dip around there. Experiment to see if this makes your voice sound more Sovereigny.

9. Distortion - very important. SF has this chart-based again, very easy - you just adjust the output volume (original and modified) and then I grab the middle of the line and drop it 3-5 db, then repeat the process increasing that point to 3-5 db above baseline. The distortion makes the voice sound more artificial without being too obvious. I used a slew rate of 100 and a low-pass around 4500Hz.

10. Repeat as necessary - many of these steps will need to be done more than once to get where you want.

11. Almost forgot - time shift. I found that speaking slowly and deliberately is better for an even voice, and then I speed it up a few percent at the end and it seems to sound more mechanical.

And take notes! Makes it much easier the next time around.

Unfortunately I don't know how to automate this, and that would depend on the software you're using anyway.

Hopefully a few people will read this and we'll see a few more Sovereign improv videos on YouTube.

It would be an awfully complex but awesome real-time voice changer for a costume.

[Wise Men]

The Reaper voice is spot on.